Tire Pressure Take-Home

un-meeting_uphill

What is the “correct” tire pressure for your bike? The simple answer is: Whatever feels right to you. Confused? Here is how it works:

In the past, many riders inflated their tires to the maximum pressure rating. Now most cyclists now recognize that the optimum pressure often is much lower.

But what is the right tire pressure? At Bicycle Quarterly, we’ve done a lot of research into the rolling resistance of tires at various pressures, and on various road surfaces.

BertoPressChart

Frank Berto’s tire pressure chart (above), first published in Bicycle Quarterly many years ago, has received much attention. (Note that the weights are per wheel, not for the entire bike.)

Berto made the chart in the 1990s, when tires were much narrower. Hardly anybody today still rides on 20 mm tires, and even 23 mm are on their way out! At the other end, 37 mm no longer is huge, as many of us ride 42 mm tires on pavement, and even wider ones on gravel. How does it all translate into the modern world?

tire_drop

Much of it depends on the tires you run. Berto measured the tire drop (above; how much the tire deflects for a given load and pressure) for dozens of tires. He then averaged the values, and drew his chart for a tire drop of 15%.

The 15% as desirable tire drop was based on the recommendations of several tire manufacturers, but not on actual testing. So the chart shows how much you need to inflate an average 1990s tire to achieve a tire drop of 15% – nothing less and nothing more.

A few years ago, Berto sent me all his original data. Looking over his measurements, it’s clear that supple tires – back then pretty much only the Michelin Hi-Lite – deflect much more than stiff ones, at the same pressure. This means that specific tires can vary quite a bit from the averages shown in the chart.

golden_gardens

To get the same tire drop with supple tires, you would need to run them at higher pressures. But is 15% tire drop really what you want with supple tires?

The answer is “No”. The 15% tire drop is an arbitrary value. However, even if it’s only by coincidence, the values in Berto’s chart actually work quite well for Compass tires. They’ll result in more than 15% tire drop, but that is OK: Comfort and speed are optimized. And that is what really matters.

track_tire_test

The biggest surprise of all our testing (above) was this: For supple tires, pressure makes little difference in performance. We tested three Vittoria tires (Rubino, CX clincher, CX tubular; all 25 mm wide) and found that the supple CX models roll as fast at 70 psi as they do at 130 psi. (For the rest of the world, that is 5 bar and 9 bar.)

The reason is simple: Higher pressure decreases the energy required to flex the tire. Less energy is lost due to internal deformation (hysteresis). But higher pressure increases the losses due to the vibrations of bike and rider. More energy goes to suspension losses. The two effects cancel each other. Whether you pump up your supple tires super-hard or ride them squishy-soft, they have the same resistance.

On the other hand, truly stiff tires feel sluggish at 15% tire drop. The stiff tire is much harder to flex, so it’s useful to minimize that flex by increasing the pressure. For stiff tires, the suspension losses do not vary as much with pressure – they’re always high – since the stiff casing transmits a lot of vibration at any pressure.

Recently, Velo-News confirmed our results: The performance of a hand-made tire with cotton casing did not change at different tire pressures. And a stiffer tire rolled slower at lower pressures than at higher ones. (It’s nice to see that our results, after having been highly controversial for years, now are becoming generally accepted.)

It can be hard to believe this, because higher pressure feels faster. Here is why: When you go faster, your bike hits more road irregularities per second: The road buzz increases in frequency. Most cyclists know: higher speed = higher frequency.

Higher tire pressure cheats you into thinking that you are going faster, because it also increases the frequency of the vibrations: higher pressure = higher frequency.

It’s natural to assume that this means: higher pressure = higher frequency = higher speed, but that is incorrect. Instead, you are looking at two different mechanisms that both increase the frequency of the road buzz.

Even after years of riding supple, wide tires, this “placebo” effect sometimes plays tricks on me. A supple tire absorbs vibrations better, so it can feel slower – until you look at your speedometer.

hahn_un-meeting

What does it all mean? Here is the take-home summary:

  • Stiff casings always will be slow. They are even slower at lower pressures.
  • Supple casings are fast, and pressure doesn’t matter.
  • On smooth roads, tire pressure is a matter of personal preference (at least with supple tires). High and low pressures offer the same performance.
  • On rough roads, lower pressures are faster. So if you want to optimize your speed on all roads, including rough ones, go with a relatively low, but safe, pressure.
  • Your tire pressure needs to be high enough to avoid pinch flats. If you get pinch flats, increase your tire pressure, or better, choose wider tires. Pinch flats are rare with wide tires.
  • On pavement, your pressure needs to be high enough that the tire does not collapse during hard cornering.
  • The minimum safe pressure is higher for more supple casings. Stiff casings hold up the bike more, and thus require less air pressure.
  • On gravel, you can run lower pressures than on pavement. On loose surfaces, the tires don’t collapse as easily, because the cornering forces are much lower.
  • Don’t run your tires so low that the casing cords start to break. That happens only at very low pressures, but if you start seeing multiple lines across the casing where cords have broken, inflate the tires a bit more.
  • Berto’s chart still is a good starting point. Inflate your tires to the pressures it recommends, then experiment by adding or letting out some air.
  • See what feels best to you. That is the optimum tire pressure for you. Don’t worry about tire pressure any further! At least on paved roads, you won’t go faster or slower if you change your tire pressure.

Even simpler, here is a summary in two sentences:

  • Ride the tire pressure that feels good to you.
  • When in doubt, let out some air.

It’s really that simple!

Further reading:

About Jan Heine, Editor, Bicycle Quarterly

Spirited rides that zig-zag across mountain ranges. Bicycle Quarterly magazine and its sister company, Compass Cycles, that turns our research into high-performance components for real-world riders.
This entry was posted in Testing and Tech, Tires. Bookmark the permalink.

46 Responses to Tire Pressure Take-Home

  1. Stephen Poole says:

    Brilliant, Jan! This is the most useful, helpful and succinct thing I’ve ever seen on this subject. Your public service medal should be on its way any day now.

  2. shastatour says:

    I would suggest a third sentence to the summary: Ride on supple tires.

  3. Robert says:

    Supple tires are nice, but a puncture renders them unusable. With stiffer sidewalls I can at least limp home to fix any flat (but I am a short-rides-only-because-I-have-to cyclist).

    • Even when I rode utility tires, I’ve never encountered a tire that could be ridden without air. Even under my light weight as a student, they compressed all the way when I had flats. (That is why it’s called a “flat” tire.) Are the new generation “puncture-resistant” tires really so stiff that you can ride them airless?

      We did test actual airless tires for rolling resistance. Even on a super-smooth track, they had 50% more resistance than the worst air-filled tire. Where I was pedaling at moderate effort on the best tires to maintain the speed we had selected for our testing, I had to work all-out on the airless tires. And the cornering, even in the wide corners of the inner apron of the track we used (not on the banking) was awful!

      Airless tires haven’t caught on because they ride so horribly. I would think that a tire that can be ridden without air would be close in feel…

      On the other hand, if you ride half a mile to a train station every morning, like many Europeans do, a flat means you missing your train and being an hour late to work. In that scenario, riding the most puncture-resistant tire makes sense.

  4. Geoff Hazel says:

    Sounds like a strategy when using supple tires is: Inflate to chart pressure. Ride. Don’t look at pressure or re-inflate until tires either (1) don’t feel right or (2) cords start showing or (3) you get pinch flats. If any of those happen, add 5-10 lbs of air and note that as the pressure you should maintain going forward.

  5. Eddie says:

    I have to ask, what about in something as specific as CX races, where most riders tend to have pressures of just 20-30 psi and are always flirting with pinching, burping, or rolling tires? Is there any takeaway from your research for such a situation or is this a similar case of trial and error (“how low can you go”)?

    • Cyclocross is a very special application. With the extra-supple FMB tubulars I run, I use the same approach – drop the pressure as much as I can until the tires start collapsing under hard cornering. That point depends on how much traction the course offers. If it’s very muddy, the pressure is lower. If there are significant paved sections, it has to be higher.

      I haven’t raced ‘cross on clinchers, but I am sure somebody else will weigh in on this.

      • Bryan Willman says:

        I have raced cross (poorly) on both clinchers and tubulars – not on tubeless. With clinchers and my very heavy weight, it takes a relatively high amount of air just to keep the tire on the rim (at one point nearly 50psi fro a 33mm tire.) Tubulars allow even someone as heavy as me to run rather lower pressures, low enough that sometimes it seems faster to add air – if there is a particular kind of hard pack, or pavement. For most PNW ‘cross courses, it seems I did best when the pressure was “just enough” to keep the tire from “slewing” sideways – it didn’t seem likely to roll off, but was no longer retaining a “tire shape”. Where that is would depend on rider weight and the tire as well as the terrain. I am quite heavy (call it 230# race weight) and went to dugast cotton tublar cross tires a few years ago and have stuck with them – because they are very compliant, and being tubulars tolerate low pressures under my weight. I don’t/wouldn’t hassle with tubulars outside of ‘cross racing.

  6. Luis Bernhardt says:

    Coming from a track racing background, where we’re used to inflating 21mm tubulars to 180 psi, I am finally convinced. I’ve been running the same 21mm tubulars in Burnaby at 110-120 psi and there’s no difference in speed or effort to hold a given speed, but the tires last longer because the casings are not stressed. (And the tread just lasts forever on a wooden track!)
    The latest bike I had built actually has room for a 32mm tire in back, and I’m slowly stockpiling and aging supple 25s and 28s as I gradually wear thru my stock of 23mm tires (it was a 3-year supply – I find that rubber aged for one or two years does appear to resist punctures better than fresh rubber; perhaps that myth will be the next object of your inquiries!).
    However, heeding some of the pioneers of the “less pressure” school of thought (Uncle Al comes to mind), I tried lowering the pressure in my 23mm tires from 120 psi to 110 psi. And promptly started getting pinch flats (btw, I weigh 175-180 lbs). I’m running Vredestein Fortezza Tri-Comps, and I’m not sure if they’d be considered a “supple” tire, as one of the big reasons I use them is because the sidewalls are not prone to failure, as on Conti GP’s and similar “skinwall” ciinchers. Tri-Comp sidewalls are definitely not as supple as Contis, but they last much longer, usually 10,000 km on the front and 5,000 on the rear before worn to the cord.
    So my question is this: why is it that “pinch flats are rare with wide tires?” It would seem to me that as long as the tube can get caught between the rim and a rock, you’ll pinch flat.
    And while we’re on the subject of tire pressures, I would like to question the convention of airing the front tire less than the rear. The accepted practice is derived from the bike’s static weight distribution with rider. I air my tires for “worst case,” 120 lbs front AND rear. This is because under braking, or when climbing out of the saddle, the front tire will have more weight on it than the rear. Worst case.

    • huges84 says:

      Pressure equals force divided by area. Which means force equals pressure times area. You hit a bump with the same force regardless of tire width. Let’s assume you run two different width tires at the same pressure. With a skinny tire a large impact force is applied to a small area, so the opposing force from the tire pressure is small. With a wide tire the impact force is applied to a large area, so the tire pressure provides a larger opposing force. But in real life we don’t run the same pressure on both tires. The wider tire will be at a lower pressure. So maybe the lower pressure will have more of an impact than the larger area. Then why do we say wider tires

      Well, wider tires are also taller than skinnier tires. So there is effectively more travel to the suspension of the tire. And as the tire gets compressed, the internal pressure is rising and more of the tire contracts the bump. Also the force needed to deflect the sidewall increases with further compression. So the force the tire applies to oppose the bump increases with increased deflection off the tire. So it takes a lot more force to overcome the travel of the tire suspension on a wider tire. Pinch flats can only happen when you bottom out the tire against the rim.

      Keep in mind the area I am talking about is the contract area with the bump. The bumps that tend to cause pinch flats are usually wider than the tire. Since wider tires have wider and shorter contact areas compared to skinny tires, the contract area of the tire and bump will always be bigger for a wider tire.

      • John Duval says:

        Using the section of a torus method, both in a mathematical (Excel spreadsheet) model and CAD model, turned up some interesting findings:

        1. the contact patch is only shorter and wider when the two sizes are inflated to the same pressure, and then only barely. Fact is, the wheel radius is so large compared to the with, that the length of the contact patch changes only very slightly.
        2. even at the same pressure, the taller tire has more “suspension travel” (which is the question that kicked off this study). When supporting a given load on a flat surface, a tire twice as wide must compress twice the distance to create the same size contact patch. So in theory, a wide tire will ride softer even at the same pressure.
        3. The wide tire deforms more, not less, than a skinny tire at either the same or proportional pressure. So lower rolling resistance does not come, even partially, from lower hysteresis.

        The pressure curve determined by Burto is indeed linear, and the difference between it and the torus section method is within 5psi across the chart. The section of a torus method is obviously still an approximation, and tire deformation is much more nuanced, but the numbers line up well enough to refine a number of hypothesis.

      • Those are interesting findings, but before we accept them, we must validate the model used to derive them. It would be interesting to actually measure the contact patch of various tires at different pressures. I can think of a few ways in which this could be done.

    • why is it that “pinch flats are rare with wide tires?”

      When you look at the tire drop curves in Berto’s chart, you see that wider tires have a much shallower slope. What this means is that you can add significant weight with less deformation than with narrow tires. Or in other words, a bump also will deform the tire much less.

      • Jason Miles says:

        Do you have analytical model for Berto’s chart? I’m guessing that would either be related to calculating the surface area of a partial torus or the surface area of a ellipse. The math geek in me cannot accept that this chart is truly linear.

        Also which tire do you think would loose air faster a large tire at low pressures or a small tire at high pressures? The smaller tire has to contain higher pressures, but it also has a smaller surface area to bleed from. I’m guessing either way the wide tires would be better because of the lower slope mentioned and the fact that they have higher volume so you will have to bleed more area to affect the pressure the same amount. I ask because I often get pinch flats when I am too lazy to pump up my tires and am riding at a too low of pressure.

      • Berto’s chart is purely empirical. He measured the actual tire drop with a lot of tires and averaged the results. Then he plotted them on the chart.

      • Jason Miles says:

        “What this means is that you can add significant weight with less deformation than with narrow tires.” I am not sure if this is correct. I think the chart shows us that with wider tires you can add significant weight without having to change your tire pressure as much.

        In my experience your other post is correct. Lower pressure higher volume tires bounce at lower frequency but higher amplitude. When riding a fatbike on large bumps you can really feel this amplitude because the undamped suspension from the tires is enough to seemingly try and buck you from the bike.

        In general I consider this the lower sloped pressure change line for larger tires to be a negative trait because it makes your pressure adjustments more sensitive.

      • I think the chart shows us that with wider tires you can add significant weight without having to change your tire pressure as much.

        The two are directly related. With wider tires, you don’t have to change your tire pressure much as you add weight, because the tire deformation doesn’t change as much.

        In general I consider this the lower sloped pressure change line for larger tires to be a negative trait because it makes your pressure adjustments more sensitive.

        That would be true if you wanted to get exactly 15% tire drop. But as the article explains, the 15% value is just a good starting point, not important by itself. With wide tires, tire pressure actually matters less, because you aren’t going to get pinch flats. Whether you have 15% or 20% tire drop (or even more) doesn’t affect the performance of wide tires.

        In a real-world example, I can ride my 42 mm tires at 55 psi or at 40 psi, or anywhere in between. They feel great and roll fast at all those pressures.

      • Jason Miles says:

        I gave an analytical derivation a try, unfortunately either my assumptions are incorrect or I screwed up the derivation because when I plug in real numbers I don’t get the correct result. If anyone else would like to give it a try, here is my attempt:

        https://goo.gl/F2I2Vf

  7. Tran says:

    I think the article meant to say that we should ride the tire pressure that feels good while observing the recommended minimum pressure, if indicated ? For the Elk Pass, that minimum is 4.5 Bars, which is more that I would like to run it at.

    • On most tires, you can ignore the minimum tire pressure. It was put on the tires for liability protection, not for technical reasons.

      Don’t worry about going lower than 4.5 bars on the Elk Pass. For most of our tires, we don’t indicate a minimum tire pressure for that reason.

  8. Ed Bernasky says:

    Wonderful Blog post!!

    Aside from cost constraints, I can’t see a reason not to run high performance, widish, supple tires because they are not only more comfortable but the time savings on a 1200 Km range from 2 hours for average tires to 4 hours savings when moving from slower tires. I don’t get many flats and the ten minutes to leisurely fix one is a small price to pay from my perspective. The additional time when running “slow” tires could be the difference between Hors Delai or successfully finishing on whatever time scale is desired.

    On a fast Brevet last year I was struggling to hold the pace or at least until we hit a 6-8 mile stretch of the nastiest chip seal and pot holed road imaginable and suddenly my required powered dropped significantly while my former “Tormentor in Chief” struggled on his 23 mm tires probably pumped to 115 psi. I was running the 32 mm Compass EL tires at 65 psi front and 80 psi rear. What I like about the Compass tires compared to the very few other performance oriented wide tires is their relative durability. I’m currently experimenting with some tubeless options from other Mfgs (Schwalbe and Specialized) but the durability is probably going to be unacceptable. Most of my rides are on Compass tires but I also use a few others.

  9. Steven Krusemark says:

    We rode our tandem around the world on specialized 27×1-3/8 tires until our supply (and money) started running low then we ran many 1-¼ tires that we picked up along the way. (It was amazing how many free tires we received from shops in the U.S. because their customers just wanted new tires) We ran the tires at ~100psi (or as high as we could go using our Zefal). We now riding the Grand Bois Hetre 650x42B tires at about 65psi. The tires ride like being on clouds and we find we go as fast than on the stiffer tires. These tires would have been nice on the rough roads of India or the sandy roads going over the Snowy Mountains in Australia! Although worldwide availability of 650B tires are limited, 27 inch tires were also not available in many places when we traveled in the 80’s. We entered India with six tires folded up, strapped to various places on the bike. Besides being more comfy, these tires fold easier and are much easier to inflate after a flat on the road with the hand pump after 60 miles of riding in 120 degree heat dodging elephants and TATA Lories!

  10. Harald says:

    Higher pressure decreases the energy required to flex the tire. Less energy is lost due to internal deformation (hysteresis). But higher pressure increases the losses due to the vibrations of bike and rider. More energy goes to suspension losses. The two effects cancel each other. Whether you pump up your supple tires super-hard or ride them squishy-soft, they have the same resistance.
    Recently, Velo-News confirmed our results: The performance of a hand-made tire with cotton casing did not change at different tire pressures.
    Isn’t this contradictory? The Velo News test was on a drum and therefore didn’t measure suspension losses. Nonetheless, rolling resistance remained approximately the same at all pressures for the supple tire. Hence, in a real-world setting, riding at lower pressures should decrease overall resistance. So I don’t think that the Velo News drum test and your real-world testing actually have the same result.

    • I was wondering about that, too. In the past, the same tests showed that higher pressures rolled faster. I assume that they use a shock absorber to keep the tire from bouncing on the patterned treads they now use on their rollers. That shock absorber also will use energy as it absorbs the shocks. So if you calibrate it correctly, you could conceivably measure suspension losses. You’d still need to validate your calibration with real-road testing…

      Real road testing has the advantage that your model is perfect. (You are testing under the actual conditions that you encounter when riding.) The key, however, is to reduce the noise in the data. You need to set up your test very, very carefully, test only when there is no wind and constant temperature, and then do a rigorous statistical analysis to show that you are measuring real differences between tires and pressures, and not just noise in the data.

  11. Gunther says:

    Certainly, higher pressure increases vibrations in the rider. But one may still discuss how this is distributed over amplitude and frequency. Maybe you are mainly hit harder (amplitude) rather than more often (frequency).

    • The frequency also changes… Pick up one end of the bike and let it drop. The bouncing you get depends on the tire pressure. Low pressure, it bounces with low frequency (but higher amplitude, since the energy you put in is the same). At higher pressure, it bounces much faster, but not so high.

      We used to do this to check our tire pressure when we were poor students and didn’t have floor pumps with gauges.

  12. Larry says:

    Since much of this discussion is predicated on how supple ones tires are …. Is there a definitive definition of a Supple tire ?

    • Defining a supple tire isn’t easy, because like so many qualities, simple measurements don’t tell the story. I wrote about that in this post.

      All I can add is that “When you ride a set of supple tires, you’ll know.”

      • Pano G. says:

        Another take-home point that may need to be highlighted is that the type and thickness of the tread compound used affects the “suppleness” of the particular tire and could be detrimental to rolling resistance by counteracting against the desirable effects of the supple sidewalls.

      • Absolutely. Designing a great tire must take everything in consideration. If the tread is too thick, or there is too much rubber on the sidewalls, or you put puncture-resistant belts under the tread – all deteriorate both comfort and performance. Fortunately, modern rubber compounds are so good that you can get great grip and great longevity without needing excessively thick tread.

        Wider tires last longer anyhow, since the wear is spread over a larger surface area. That means that modern supple tires may be expensive when you buy them, but they don’t cost much per mile.

  13. Arthur says:

    Are there any pictures that show what broken casing cords look like?

  14. PabloB says:

    As the road is very rough in my area, so I prefer to ride at very slow pressures too, but my rear tire has often collapsed during cornering. Thou I have limited wheels to compare this effect seems to more prone when same tires are mounted in (most road) narrow rims than in wider (like the Rhino Lite for example).

    Any Thoughts

  15. alliwant says:

    Something about this statement seems a little off: “Higher pressure decreases the energy required to flex the tire. Less energy is lost due to internal deformation (hysteresis).” Isn’t is more accurate to say that higher pressure *increases* the energy required to flex the tire, which would result in less internal deformation (and by extension, moreenergy transmitted by vibration)?

  16. Michael says:

    Speaking of tire deflections…
    I ride through a gated community and the track that the gate glides on is a metal strip with a tongue about two inches high and one deep that runs the whole length of the track.. So its like someone riding over a spike strip without the spikes. Just a blunt ridge.
    Is repeated riding over these types of ridges bad for the tire, since it is such a directed point of force into the tire as the tire rides over the tongue?

    Haven’t gotten any pinch flats or anything. Wondering if it stresses the casing or something. No way around the ridges except hopping off and walking the bike across.

    I ride 650b cypres and loups at 45 psi.

  17. thebvo says:

    I remember switching to 48mm wide utility tires and they felt so much better than the cheap 28’s I had been riding previously. But then I rode supple tires and even at the skinny ish size of 32 it was totally noticeable. It’s not my metaphor but supple tires truly make roads feel freshly paved. And now we can have wide AND supple (fast) tires!!!
    Thanks for pushing the envelope in this direction and making it fun and easy to understand for fools like me.

  18. Bill Wood says:

    Hi Jan, I have some hard to interpret results here. Inspired by the Berto chart, I measured tire drop for two sets of tires, front and rear. I measured unweighted and weighted (70 lbs front, 130 lbs rear) rollout at 3 different tire pressures. I measured along a tape measure for 5 rotations, then divided the difference between weighted and unweighted circumference by the measured tire width. Each measurement was repeated several times and the average taken. Measurements were consistent. These graphs show the %tire drop vs PSI:

    Roubaix Pro 30/32 (measured @33mm): https://docs.google.com/spreadsheets/d/1ptSMWTu4KUs8ywOknlK5mwyny-Qa2PO2aq5oD9qastc/pubchart?oid=286831095&format=interactive

    Continental GP4000S II 28 (measured @31mm):
    https://docs.google.com/spreadsheets/d/1ptSMWTu4KUs8ywOknlK5mwyny-Qa2PO2aq5oD9qastc/pubchart?oid=258933175&format=interactive

    As you can see, the Contis, while narrower and more supple, require substantially less tire pressure to provide a 15% drop (65psi vs 90psi on the rear wheel and 33psi vs 35psi on the front), which is opposite what I expected based on your post.

    Thoughts?

    PS. Soon I will measure Bon Jon Pass 35 extra lights.

    • Do we know that the Contis are more supple? In my experience, Contis usually ride quite harshly… I’ll be interested to see more results.

      • Bill Wood says:

        The Contis certainly feel more supple, and are more collapsed when mounted without air on the rim, kind of like Compass tires, whereas the Roubaix Pros sit tall. Also the Conti 28s have great reviews for rolling resistance. I will be trying them at lower pressure (50 front 75 rear), see how they feel.

    • Bill Wood says:

      Oops I meant to say “divided the difference between weighted and unweighted RADIUS by the measured tire width”

  19. Another Frank says:

    One reader asked for a picture of broken casing cords. Perhaps like this? Continental GP4000S II in my experience have fragile sidewalls: https://dl.dropboxusercontent.com/u/9474052/GP4000S%20II%2028mm.JPG

Comments are closed.